In [1]:
from util import *
from glob import glob
import matplotlib.pyplot as plt
from shapely import wkt
In [2]:
df = load_AOIs()
df
Out[2]:
Taranaki AOI SSP 4.5 (p50) SSP 4.5 (p83) SSP 8.5 (p50) SSP 8.5 (p83) Rate SSP 4.5 (p50) Rate SSP 4.5 (p83) Rate SSP 8.5 (p50) Rate SSP 8.5 (p83) match match_score
7 NORTH TongaporutuRiver 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 TongapurutuRiverCliffs 93.750000
11 SOUTH HangatahuaRiver_South 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 HangatahuRiver_South 97.435897
21 SOUTH Rahotu 0.58 0.78 0.84 1.10 0.0058 0.0078 0.0084 0.0110 Rahotu 100.000000
20 SOUTH Pihama 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 Pihama 100.000000
19 SOUTH OpunakeBeach 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 OpunakeBeachCliffs 100.000000
18 SOUTH OhaweBeach 0.57 0.78 0.83 1.10 0.0057 0.0078 0.0083 0.0110 OhaweBeach 100.000000
17 SOUTH Oeo 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 Oeo 100.000000
16 SOUTH Manutahi 0.57 0.78 0.83 1.10 0.0057 0.0078 0.0083 0.0110 Manutahi 100.000000
15 SOUTH ManaBay 0.57 0.78 0.83 1.10 0.0057 0.0078 0.0083 0.0110 ManaBayCliffs 100.000000
14 SOUTH KaupokonuiBeach 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 KaupokonuiBeach 100.000000
13 SOUTH Kakaramea 0.57 0.78 0.83 1.10 0.0057 0.0078 0.0083 0.0110 Kakaramea 100.000000
12 SOUTH Hawera_WaihiBeach 0.57 0.78 0.83 1.10 0.0057 0.0078 0.0083 0.0110 Hawera_WaihiBeach 100.000000
0 NORTH Mohakatino_River 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 MohakatinoRiver 100.000000
10 SOUTH CapeEgmont 0.58 0.78 0.84 1.10 0.0058 0.0078 0.0084 0.0110 CapeEgmont 100.000000
9 NORTH Waitara 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 Waitara 100.000000
8 NORTH UrenuiRiver_North 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 UrenuiRiverNorthCliffs 100.000000
6 NORTH PariokariwaPoint 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 PariokariwaPointCliffs 100.000000
5 NORTH Onaero 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 OnaeroCliff 100.000000
4 NORTH Oakura_South 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 OakuraSouth 100.000000
3 NORTH Oakura 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 Oakura 100.000000
2 NORTH NewPlymouth_Airport 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 NewPlymouthAirport 100.000000
1 NORTH NewPlymouth 0.57 0.78 0.84 1.10 0.0057 0.0078 0.0084 0.0110 NewPlymouthCliffs 100.000000
22 SOUTH WainuiBeach 0.57 0.78 0.83 1.09 0.0057 0.0078 0.0083 0.0109 WainuiBeach 100.000000
23 SOUTH WaipipiBeach 0.57 0.78 0.83 1.09 0.0057 0.0078 0.0083 0.0109 WaipipiBeachCliffs 100.000000
In [3]:
site = df.match.sample(1).iloc[0]
site = "Oeo"
site
Out[3]:
'Oeo'
In [4]:
gdf = gpd.read_file(f"Shapefiles/{site}_intersects.shp")
gdf.crs = 2193
gdf = enrich_df(gdf)
transect_metadata = get_transect_metadata(f"Shapefiles/{site}_TransectLines.shp")
if site == "ManaBayCliffs":
  print("Flipping")
  for k, v in transect_metadata.items():
    transect_metadata[k]["Azimuth"] = v["Azimuth"] + 180
In [5]:
group_data = gdf[gdf.TransectID == 436]
group_data
Out[5]:
OBJECTID TransectID TransOrder BaselineID ShorelineI Distance IntersectX IntersectY Uncertaint AOI geometry Date Year YearsSinceBase YearsUntilFuture
1337 1624 436 436 1 31/10/2016 15.368598 1.684813e+06 5.621396e+06 2.091626 Oeo POINT (1684812.590 5621396.412) 2016-10-31 2016 76.832307 83.167693
1338 1625 436 436 1 18/03/2022 8.864836 1.684819e+06 5.621399e+06 2.429774 Oeo POINT (1684818.570 5621398.970) 2022-03-18 2022 82.209446 77.790554
In [6]:
linear_model = LinearRegression().fit(
    group_data.YearsSinceBase.to_numpy().reshape(-1, 1), group_data.Distance
)
In [7]:
linear_model.coef_[0], linear_model.intercept_
Out[7]:
(-1.2095208005543252, 108.29887058148721)
In [8]:
linear_model.score(group_data.YearsSinceBase.to_numpy().reshape(-1, 1), group_data.Distance)
Out[8]:
1.0
In [9]:
linear_model.coef_[0] * 160 + linear_model.intercept_
Out[9]:
-85.22445750720483
In [10]:
linear_model.predict([[160]])
Out[10]:
array([-85.22445751])
In [11]:
ax = group_data.plot(x="YearsSinceBase", y="Distance", kind="scatter")
years_to_predict = np.array([0] + group_data.YearsSinceBase.tolist() + [160])
print(years_to_predict)
ax.plot(
    years_to_predict,
    linear_model.predict(years_to_predict.reshape(-1, 1)),
    color="red",
)

[  0.          76.83230664  82.20944559 160.        ]
Out[11]:
[<matplotlib.lines.Line2D at 0x7f859409eef0>]
No description has been provided for this image
In [12]:
latest_row = group_data[group_data.Date == group_data["Date"].max()].iloc[0]
latest_row
Out[12]:
OBJECTID                                                    1625
TransectID                                                   436
TransOrder                                                   436
BaselineID                                                     1
ShorelineI                                            18/03/2022
Distance                                                8.864836
IntersectX                                         1684818.57004
IntersectY                                         5621398.96952
Uncertaint                                              2.429774
AOI                                                          Oeo
geometry            POINT (1684818.5700383317 5621398.969516366)
Date                                                  2022-03-18
Year                                                        2022
YearsSinceBase                                         82.209446
YearsUntilFuture                                       77.790554
Name: 1338, dtype: object
In [13]:
pred = linear_model.predict([[latest_row.YearsSinceBase]])[0]
print(f"Linear model predicts {round(latest_row.Distance - pred)}m erosion")

Linear model predicts 0m erosion
In [14]:
linear_models = fit(gdf, transect_metadata)
linear_models.loc[linear_models.slope > 0, "slope"] = pd.NA
rolled_slopes = linear_models.groupby("group").slope.rolling(10, min_periods=1).mean().dropna().reset_index(level=0)
linear_models.slope = rolled_slopes.slope
linear_models.dropna(inplace=True)
linear_models
Out[14]:
TransectID slope intercept group r2_score mae mse rmse
0 2 -0.025895 11.839523 0 1.000000 0.000000e+00 0.000000e+00 0.000000e+00
1 3 -0.026103 11.394666 0 1.000000 1.776357e-15 3.155444e-30 1.776357e-15
2 4 -0.032111 12.067457 0 1.000000 0.000000e+00 0.000000e+00 0.000000e+00
3 5 -0.037306 12.952147 0 1.000000 0.000000e+00 0.000000e+00 0.000000e+00
4 6 -0.041094 14.552892 0 1.000000 0.000000e+00 0.000000e+00 0.000000e+00
... ... ... ... ... ... ... ... ...
665 806 -0.088458 17.144344 50 0.462361 2.814825e+00 9.597485e+00 3.097981e+00
666 807 -0.098385 17.888823 50 0.439388 3.502577e+00 1.287081e+01 3.587592e+00
667 808 -0.100127 16.604210 50 0.510860 1.851194e+00 6.213661e+00 2.492722e+00
668 809 -0.105771 5.736368 50 0.130797 1.977019e+00 5.294052e+00 2.300881e+00
669 810 -0.105771 4.466733 50 0.167198 1.881087e+00 4.792745e+00 2.189234e+00

657 rows × 8 columns

In [15]:
results = predict(gdf, linear_models, transect_metadata)
results
Out[15]:
TransectID slope intercept group r2_score mae mse rmse BaselineID Year ... linear_model_distance sqrt_model_point sqrt_model_predicted_distance sqrt_model_distance BH_model_point BH_model_predicted_distance BH_model_distance Sunamura_model_point Sunamura_model_predicted_distance Sunamura_model_distance
0 2.0 -0.025895 11.839523 0.0 1.000000 0.000000e+00 0.000000e+00 0.000000e+00 1 2100 ... 2.014361 POINT (1688035.5534006841 5620798.3280850835) 2.575163 7.135573 POINT (1688043.1672309272 5620827.768582187) -27.833936 37.544672 POINT (1688037.0176938204 5620803.9900865285) -3.273120 12.983856
1 3.0 -0.026103 11.394666 0.0 1.000000 1.776357e-15 3.155444e-30 1.776357e-15 1 2100 ... 2.013439 POINT (1688045.4118067573 5620796.610869867) 2.055679 7.175904 POINT (1688053.2101313008 5620825.9603533875) -28.312167 37.543750 POINT (1688046.9030184844 5620802.223138728) -3.751323 12.982906
2 4.0 -0.032111 12.067457 0.0 1.000000 0.000000e+00 0.000000e+00 0.000000e+00 1 2100 ... 1.510104 POINT (1688055.3321562614 5620795.049295843) 0.578900 7.860823 POINT (1688062.928654537 5620823.222719841) -28.600693 37.040415 POINT (1688056.5343737882 5620799.508006012) -4.039046 12.478768
3 5.0 -0.037306 12.952147 0.0 1.000000 0.000000e+00 0.000000e+00 0.000000e+00 1 2100 ... 1.620997 POINT (1688065.1311122181 5620792.999345171) -0.394737 8.998973 POINT (1688072.6592310544 5620820.126478494) -28.547072 37.151308 POINT (1688066.0911000874 5620796.458604978) -3.984731 12.588967
4 6.0 -0.041094 14.552892 0.0 1.000000 0.000000e+00 0.000000e+00 0.000000e+00 1 2100 ... 1.950942 POINT (1688074.7352682156 5620790.222629285) -0.149400 10.078169 POINT (1688081.7768776012 5620816.705543141) -27.552484 37.481253 POINT (1688075.4651071057 5620792.967493336) -2.989637 12.918406
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
652 806.0 -0.088458 17.144344 50.0 0.462361 2.814825e+00 9.597485e+00 3.097981e+00 1 2100 ... 2.020175 POINT (1687892.4113368606 5620820.913072934) -14.503338 19.514583 POINT (1687895.824934831 5620838.622988925) -32.539240 37.550485 POINT (1687891.1748046514 5620814.497882353) -7.970063 12.981308
653 807.0 -0.098385 17.888823 50.0 0.439388 3.502577e+00 1.287081e+01 3.587592e+00 1 2100 ... 2.435977 POINT (1687902.105572404 5620821.720250634) -17.310369 21.893612 POINT (1687904.5866393007 5620837.600276526) -33.383045 37.966288 POINT (1687900.7938008038 5620813.324279123) -8.812541 13.395784
654 808.0 -0.100127 16.604210 50.0 0.510860 1.851194e+00 6.213661e+00 2.492722e+00 1 2100 ... 5.662195 POINT (1687910.64721948 5620822.479462039) -19.218452 25.464476 POINT (1687911.7553046227 5620838.168409724) -34.946482 41.192506 POINT (1687910.0242252424 5620813.6587283) -10.375745 16.621769
655 809.0 -0.105771 5.736368 50.0 0.130797 1.977019e+00 5.294052e+00 2.300881e+00 1 2100 ... 17.597616 POINT (1687920.859272824 5620834.459251945) -32.105525 38.516099 POINT (1687921.5934341024 5620849.052623445) -46.717352 53.127926 POINT (1687920.3588562713 5620824.512166921) -22.145860 28.556435
656 810.0 -0.105771 4.466733 50.0 0.167198 1.881087e+00 4.792745e+00 2.189234e+00 1 2100 ... 18.054265 POINT (1687928.2598098365 5620836.8874109285) -33.375160 38.972749 POINT (1687927.9589022126 5620851.496139112) -47.986986 53.584576 POINT (1687928.4649134837 5620826.929858534) -23.415495 29.013085

657 rows × 23 columns

In [16]:
pd.DataFrame(results).plot(x="TransectID", y=["linear_model_distance", "sqrt_model_distance", "BH_model_distance", "Sunamura_model_distance"], figsize=(10,10))
Out[16]:
<AxesSubplot: xlabel='TransectID'>
No description has been provided for this image
In [17]:
csvs = sorted(glob("Projected_Shoreline_Polygons/*.csv"))
csvs
Out[17]:
['Projected_Shoreline_Polygons/CapeEgmont_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/CapeEgmont_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/CapeEgmont_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/CapeEgmont_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/HangatahuRiver_South_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/HangatahuRiver_South_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/HangatahuRiver_South_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/HangatahuRiver_South_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Hawera_WaihiBeach_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Hawera_WaihiBeach_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Hawera_WaihiBeach_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Hawera_WaihiBeach_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Kakaramea_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Kakaramea_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Kakaramea_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Kakaramea_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/KaupokonuiBeach_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/KaupokonuiBeach_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/KaupokonuiBeach_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/KaupokonuiBeach_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/ManaBayCliffs_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/ManaBayCliffs_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/ManaBayCliffs_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/ManaBayCliffs_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Manutahi_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Manutahi_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Manutahi_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Manutahi_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/MohakatinoRiver_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/MohakatinoRiver_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/MohakatinoRiver_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/MohakatinoRiver_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/NewPlymouthAirport_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/NewPlymouthAirport_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/NewPlymouthAirport_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/NewPlymouthAirport_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/NewPlymouthCliffs_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/NewPlymouthCliffs_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/NewPlymouthCliffs_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/NewPlymouthCliffs_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/OakuraSouth_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/OakuraSouth_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/OakuraSouth_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/OakuraSouth_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Oakura_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Oakura_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Oakura_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Oakura_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Oeo_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Oeo_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Oeo_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Oeo_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/OhaweBeach_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/OhaweBeach_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/OhaweBeach_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/OhaweBeach_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/OnaeroCliff_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/OnaeroCliff_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/OnaeroCliff_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/OnaeroCliff_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/OpunakeBeachCliffs_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/OpunakeBeachCliffs_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/OpunakeBeachCliffs_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/OpunakeBeachCliffs_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/PariokariwaPointCliffs_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/PariokariwaPointCliffs_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/PariokariwaPointCliffs_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/PariokariwaPointCliffs_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Pihama_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Pihama_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Pihama_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Pihama_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Rahotu_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Rahotu_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Rahotu_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Rahotu_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/TongapurutuRiverCliffs_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/TongapurutuRiverCliffs_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/TongapurutuRiverCliffs_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/TongapurutuRiverCliffs_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/UrenuiRiverNorthCliffs_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/UrenuiRiverNorthCliffs_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/UrenuiRiverNorthCliffs_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/UrenuiRiverNorthCliffs_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/WainuiBeach_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/WainuiBeach_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/WainuiBeach_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/WainuiBeach_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/WaipipiBeachCliffs_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/WaipipiBeachCliffs_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/WaipipiBeachCliffs_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/WaipipiBeachCliffs_Rate SSP 8.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Waitara_Rate SSP 4.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Waitara_Rate SSP 4.5 (p83)_results.csv',
 'Projected_Shoreline_Polygons/Waitara_Rate SSP 8.5 (p50)_results.csv',
 'Projected_Shoreline_Polygons/Waitara_Rate SSP 8.5 (p83)_results.csv']
In [18]:
def read(f):
    df = pd.read_csv(f)
    bits = os.path.basename(f).split("_Rate ")
    df["site"] = bits[0]
    df["SLR"] = bits[1].replace("_results.csv", "")
    return df
csv_results = pd.concat(read(f) for f in csvs)
csv_results.rename(columns={"linear_model_distance": "HistoricRate", "sqrt_model_distance": "WD", "BH_model_distance": "BH", "Sunamura_model_distance": "Sunamura"}, inplace=True)
csv_results
Out[18]:
TransectID slope intercept group r2_score mae mse rmse original_slope BaselineID ... sqrt_model_predicted_distance WD BH_model_point BH_model_predicted_distance BH Sunamura_model_point Sunamura_model_predicted_distance Sunamura site SLR
0 1.0 -0.022164 10.625580 0.0 0.411895 0.664743 0.544681 0.738025 -0.022164 1 ... 4.586685 4.013788 POINT (1665207.8750121973 5650278.888946311) -9.797479 18.397952 POINT (1665196.4019923976 5650276.774972595) 1.868672 6.731801 CapeEgmont SSP 4.5 (p50)
1 2.0 -0.022279 10.656445 0.0 0.398535 0.683677 0.610930 0.781620 -0.022395 1 ... 4.585985 4.443011 POINT (1665209.6745226474 5650269.052195227) -9.785150 18.814146 POINT (1665198.201495615 5650266.938220179) 1.881008 7.147988 CapeEgmont SSP 4.5 (p50)
2 3.0 -0.021290 10.899940 0.0 0.273105 0.701507 0.801016 0.894995 -0.019310 1 ... 5.099174 4.235516 POINT (1665211.090950729 5650259.144821454) -9.383284 18.717974 POINT (1665199.617985495 5650257.030857792) 2.282811 7.051879 CapeEgmont SSP 4.5 (p50)
3 4.0 -0.018583 11.027412 0.0 0.068164 0.948226 1.207577 1.098898 -0.010462 1 ... 5.964224 3.606545 POINT (1665212.3512731898 5650249.208730113) -8.822692 18.393461 POINT (1665200.878476966 5650247.094797593) 2.843231 6.727538 CapeEgmont SSP 4.5 (p50)
4 5.0 -0.019135 12.032552 0.0 0.330090 0.727144 0.746245 0.863855 -0.021344 1 ... 6.818871 3.052131 POINT (1665213.2492725323 5650239.271966417) -7.905925 17.776927 POINT (1665201.794168372 5650237.063987547) 3.760034 6.110968 CapeEgmont SSP 4.5 (p50)
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
791 906.0 -0.070702 5.588488 2.0 0.220615 1.327923 2.968206 1.722848 0.034744 3 ... -20.941152 29.335741 POINT (1703962.292991441 5683742.0238418225) -45.695388 54.089977 POINT (1703971.500439912 5683768.082676255) -18.057733 26.452322 Waitara SSP 8.5 (p83)
792 907.0 -0.063472 5.026776 2.0 0.301152 1.426352 2.950875 1.717811 0.042744 3 ... -18.790166 27.062807 POINT (1703952.0700438959 5683747.428869862) -45.100402 53.373042 POINT (1703962.1629671226 5683773.1565354485) -17.463834 25.736474 Waitara SSP 8.5 (p83)
793 908.0 -0.055162 4.236645 2.0 0.343306 1.432894 2.601280 1.612848 0.044202 3 ... -16.462035 23.537088 POINT (1703942.5813463528 5683753.249645808) -44.560901 51.635954 POINT (1703953.2815694462 5683778.729363978) -16.925583 24.000636 Waitara SSP 8.5 (p83)
794 909.0 -0.045386 3.897024 2.0 0.313346 1.131100 1.682665 1.297176 0.033215 3 ... -13.133217 18.881430 POINT (1703933.9814501503 5683759.456020689) -43.336295 49.084508 POINT (1703944.9163240474 5683784.834320479) -15.702447 21.450660 Waitara SSP 8.5 (p83)
795 910.0 -0.032668 3.792881 2.0 0.138292 0.630749 0.702120 0.837926 0.012724 3 ... -8.465358 12.876748 POINT (1703925.5960213046 5683766.343932675) -41.405649 45.817040 POINT (1703936.8261300405 5683791.590883901) -13.773713 18.185104 Waitara SSP 8.5 (p83)

60284 rows × 26 columns

In [19]:
most_negative = csv_results.sort_values(by="HistoricRate").iloc[0]
most_negative
Out[19]:
TransectID                                                                  101.0
slope                                                                   -0.123515
intercept                                                               29.077204
group                                                                         8.0
r2_score                                                                 0.844693
mae                                                                       2.02501
mse                                                                      5.902778
rmse                                                                     2.429563
original_slope                                                          -0.198268
BaselineID                                                                      7
Year                                                                         2100
ocean_point                          POINT (1721729.4107415285 5684661.107968306)
linear_model_point                   POINT (1721980.6946175254 5684228.838585738)
linear_model_predicted_distance                                          9.314724
HistoricRate                                                             0.000405
sqrt_model_point                     POINT (1721990.3767234418 5684212.183008734)
sqrt_model_predicted_distance                                           -9.950566
WD                                                                      19.265696
BH_model_point                       POINT (1721993.6404886057 5684206.568538858)
BH_model_predicted_distance                                            -16.444751
BH                                                                      25.759881
Sunamura_model_point                 POINT (1721984.6867233445 5684221.971192792)
Sunamura_model_predicted_distance                                          1.3713
Sunamura                                                                  7.94383
site                                                       UrenuiRiverNorthCliffs
SLR                                                                 SSP 4.5 (p83)
Name: 69, dtype: object
In [20]:
ax = group_data.plot(x="YearsSinceBase", y="Distance", kind="scatter")
x = np.array([0, 160])
ax.plot(x, most_negative.original_slope*x + most_negative.intercept, color="red", label="original slope")
ax.plot(x, most_negative.slope*x + most_negative.intercept, color="blue", label="adjusted slope")
ax.set_title(f"{most_negative.site} Transect {most_negative.TransectID}")
ax.legend()
Out[20]:
<matplotlib.legend.Legend at 0x7f85943a71f0>
No description has been provided for this image
In [21]:
most_negative.slope * 82 + most_negative.intercept
Out[21]:
18.94893317378198
In [22]:
csv_results.SLR.unique()
Out[22]:
array(['SSP 4.5 (p50)', 'SSP 4.5 (p83)', 'SSP 8.5 (p50)', 'SSP 8.5 (p83)'],
      dtype=object)
In [23]:
ymin, ymax = csv_results[["HistoricRate", "WD", "BH", "Sunamura"]].min().min(), csv_results[["HistoricRate", "WD", "BH", "Sunamura"]].max().max()
ymin, ymax
Out[23]:
(0.000405442195559, 575.5160150612829)
In [24]:
plt.style.use('ggplot')
for SLR in csv_results.SLR.unique():
    subset = csv_results[csv_results.SLR == SLR]
    ax = subset.groupby("site").boxplot(column=["HistoricRate", "WD", "BH", "Sunamura"], figsize=(20,30), showfliers=False, layout=(6,4), ylabel="meters")
    plt.ylim([ymin, ymax])
    plt.subplots_adjust(hspace=0.5)
    plt.suptitle("Taranaki Cliff Erosion projected distance at 2100, " + SLR)
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
In [25]:
plt.style.use('ggplot')
for SLR in csv_results.SLR.unique():
    subset = csv_results[csv_results.SLR == SLR]
    grouped = subset.groupby("site")
    fig, axs = plt.subplots(figsize=(20,35), 
                            nrows=6, ncols=4,
                            gridspec_kw=dict(hspace=0.5))
    targets = zip(grouped.groups.keys(), axs.flatten())
    for i, (key, ax) in enumerate(targets):
        group = grouped.get_group(key)
        ax.plot(group[["HistoricRate", "WD", "BH", "Sunamura"]], label=["HistoricRate", "WD", "BH", "Sunamura"], linestyle="", marker=".")
        ax.set_title(key)
        ax.set_ylabel("meters")
        ax.set_ylim([ymin, ymax])
    #ax.legend()
    plt.suptitle("Taranaki Cliff Erosion projected distance at 2100, " + SLR)
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
In [26]:
csv_results["lat"] = csv_results.linear_model_point.apply(lambda s: wkt.loads(s).y)
In [42]:
subset.HistoricRate.rolling(10, min_periods=1).mean()
Out[42]:
130     3.087640
129     3.265226
132     3.434248
128     3.582721
131     3.637203
         ...    
802    10.560169
803    11.437460
804    11.280412
805    11.303683
806    11.203467
Name: HistoricRate, Length: 15071, dtype: float64
In [59]:
csv_results.sort_values("lat", inplace=True)
for SLR in csv_results.SLR.unique():
    fig,ax = plt.subplots(figsize=(15,15))
    subset = csv_results[csv_results.SLR == SLR]
    for model in ["HistoricRate", "WD", "BH", "Sunamura"]:
        ax.plot(subset[model].rolling(500).mean(), subset.lat, label=model)
    ax.set_title("Taranaki Cliff Erosion projected distance at 2100, " + SLR)
    ax.legend()
    ax.set_xlabel("meters")
    ax.set_ylabel("latitude")
    ax.set_xlim(ymin, ymax)
    plt.show()
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
In [28]:
csv_results.groupby("SLR").boxplot(column=["HistoricRate", "WD", "BH", "Sunamura"], figsize=(15,15), showfliers=False, ylabel="meters")
plt.suptitle("Taranaki Cliff Erosion projected distance at 2100")
Out[28]:
Text(0.5, 0.98, 'Taranaki Cliff Erosion projected distance at 2100')
No description has been provided for this image
In [29]:
csv_results[["r2_score", "mae", "mse", "rmse"]].describe()
Out[29]:
r2_score mae mse rmse
count 6.028400e+04 60284.000000 60284.000000 60284.000000
mean 6.277615e-01 2.234403 12.755459 2.649437
std 3.385316e-01 1.999493 40.180625 2.395003
min 2.670516e-08 0.000000 0.000000 0.000000
25% 3.345064e-01 0.953863 1.235364 1.111469
50% 7.446224e-01 1.789659 4.366453 2.089606
75% 9.337973e-01 2.931538 12.176169 3.489437
max 1.000000e+00 28.657652 1349.487919 36.735377
In [30]:
gpd.GeoSeries(results.linear_model_point).distance(gpd.GeoSeries(results.Sunamura_model_point)).describe()
Out[30]:
count    657.000000
mean      10.961338
std        0.009703
min       10.924303
25%       10.957830
50%       10.964826
75%       10.967859
max       10.972655
dtype: float64
In [31]:
results.set_geometry(f"linear_model_point", inplace=True, crs=2193)
poly = prediction_results_to_polygon(results)
In [32]:
poly.explore(tiles="Esri.WorldImagery")
Out[32]:
Make this Notebook Trusted to load map: File -> Trust Notebook
In [33]:
site = df.match.sample(1).iloc[0]
site = "CapeEgmont"
print(site)
gdf = gpd.read_file(f"Shapefiles/{site}_intersects.shp")
gdf.crs = 2193
gdf = enrich_df(gdf)
transect_metadata = get_transect_metadata(f"Shapefiles/{site}_TransectLines.shp")
m = gpd.read_file(f"Shapefiles/{site}_TransectLines.shp").explore("TransectID", tiles="Esri.WorldImagery", max_zoom=22)
linear_models = fit(gdf, transect_metadata)
linear_models.loc[linear_models.slope > 0, "slope"] = pd.NA
rolled_slopes = linear_models.groupby("group").slope.rolling(10, min_periods=1).mean().dropna().reset_index(level=0)
linear_models.slope = rolled_slopes.slope
linear_models.dropna(inplace=True)
results = predict(gdf, linear_models, transect_metadata)
results["geometry"] = results["linear_model_point"]
pd.concat([gdf, results])[["Year", "geometry"]].explore("Year", m=m)
gpd.GeoSeries(results.ocean_point, crs=2193).explore(m=m, color="blue")
#gpd.GeoSeries(poly, crs=2193).explore(color="pink", m=m)
results.set_geometry(f"linear_model_point", inplace=True, crs=2193)
polygon = prediction_results_to_polygon(results)
polygon.boundary.explore(m=m)
m

CapeEgmont

/tmp/ipykernel_19942/2537168661.py:15: FutureWarning: You are adding a column named 'geometry' to a GeoDataFrame constructed without an active geometry column. Currently, this automatically sets the active geometry column to 'geometry' but in the future that will no longer happen. Instead, either provide geometry to the GeoDataFrame constructor (GeoDataFrame(... geometry=GeoSeries()) or use `set_geometry('geometry')` to explicitly set the active geometry column.
  results["geometry"] = results["linear_model_point"]
Out[33]:
Make this Notebook Trusted to load map: File -> Trust Notebook
In [34]:
run_all_parallel()

  0%|          | 0/24 [00:00<?, ?it/s]
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)

Flipping

/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
/home/nyou045/.local/lib/python3.10/site-packages/sklearn/metrics/_regression.py:918: UndefinedMetricWarning: R^2 score is not well-defined with less than two samples.
  warnings.warn(msg, UndefinedMetricWarning)
In [35]:
def read_file(f):
  df = gpd.read_file(f)
  df["filename"] = f
  return df
samples = pd.concat(read_file(f) for f in glob("Projected_Shoreline_Polygons/*.shp"))
samples.explore("filename", tiles="Esri.WorldImagery", style_kwds={"fill": False})
Out[35]:
Make this Notebook Trusted to load map: File -> Trust Notebook
In [36]:
samples
Out[36]:
FID geometry filename
0 0 POLYGON ((1715849.009 5609494.719, 1715855.318... Projected_Shoreline_Polygons/Manutahi_linear.shp
1 1 POLYGON ((1715466.059 5609873.034, 1715472.150... Projected_Shoreline_Polygons/Manutahi_linear.shp
2 2 POLYGON ((1715513.294 5609840.663, 1715520.158... Projected_Shoreline_Polygons/Manutahi_linear.shp
3 3 POLYGON ((1714209.583 5610944.439, 1714217.745... Projected_Shoreline_Polygons/Manutahi_linear.shp
0 0 POLYGON ((1738150.672 5704339.905, 1737658.096... Projected_Shoreline_Polygons/MohakatinoRiver_S...
... ... ... ...
8 8 POLYGON ((1705844.715 5615922.738, 1705849.620... Projected_Shoreline_Polygons/Hawera_WaihiBeach...
9 9 POLYGON ((1704463.125 5616726.808, 1704470.510... Projected_Shoreline_Polygons/Hawera_WaihiBeach...
10 10 POLYGON ((1705703.998 5615984.463, 1705706.712... Projected_Shoreline_Polygons/Hawera_WaihiBeach...
11 11 POLYGON ((1704072.428 5616832.182, 1704082.199... Projected_Shoreline_Polygons/Hawera_WaihiBeach...
12 12 POLYGON ((1704414.114 5616818.559, 1704286.371... Projected_Shoreline_Polygons/Hawera_WaihiBeach...

6461 rows × 3 columns

In [37]:
# Compare different SLR rates
site = df.match.sample(1).iloc[0]
print(site)
gdf = gpd.read_file(f"Shapefiles/{site}_intersects.shp")
gdf.crs = 2193
gdf = enrich_df(gdf)
transect_metadata = get_transect_metadata(f"Shapefiles/{site}_TransectLines.shp")
linear_models = fit(gdf, transect_metadata)
print(linear_models.describe())
# Decreasing distance indicates erosion. Increasing indicates accretion. Erosion only
linear_models.loc[linear_models.slope > 0, "slope"] = pd.NA
rolled_slopes = linear_models.groupby("group").slope.rolling(10, min_periods=1).mean().dropna().reset_index(level=0)
linear_models.slope = rolled_slopes.slope
linear_models.dropna(inplace=True)
print(linear_models.describe())
low_proj_slr = predict(gdf, linear_models, transect_metadata, Proj_SLR=0.007)
high_proj_slr = predict(gdf, linear_models, transect_metadata, Proj_SLR=0.015)
m = gpd.GeoSeries(low_proj_slr.sqrt_model_point, crs=2193).explore(color="blue", tiles="Esri.WorldImagery")
gpd.GeoSeries(high_proj_slr.sqrt_model_point, crs=2193).explore(color="green", m=m)

OhaweBeach
       TransectID       slope   intercept       group    r2_score         mae  \
count  855.000000  855.000000  855.000000  855.000000  855.000000  855.000000   
mean   451.877193   -0.259123   34.073809    7.836257    0.790456    1.885495   
std    259.562571    0.213989   17.044958    4.062895    0.218531    1.236279   
min      1.000000   -1.282722    8.169865    0.000000    0.000036    0.000000   
25%    230.500000   -0.310748   22.779039    6.000000    0.729918    1.118052   
50%    456.000000   -0.206993   30.100516    8.000000    0.861166    1.636407   
75%    672.000000   -0.118724   39.704068   10.500000    0.941404    2.312036   
max    898.000000    0.012665  115.700904   14.000000    1.000000    8.484189   

              mse        rmse  
count  855.000000  855.000000  
mean     6.948073    2.224217  
std     11.315229    1.415371  
min      0.000000    0.000000  
25%      1.811927    1.346074  
50%      3.897901    1.974310  
75%      7.422853    2.724490  
max     99.519188    9.975930  
       TransectID       slope   intercept       group    r2_score         mae  \
count  854.000000  854.000000  854.000000  854.000000  854.000000  854.000000   
mean   451.552693   -0.263409   34.099787    7.830211    0.791335    1.886942   
std    259.541085    0.205639   17.038002    4.061425    0.217142    1.236279   
min      1.000000   -1.282722    8.169865    0.000000    0.000036    0.000000   
25%    230.250000   -0.289099   22.784427    6.000000    0.730349    1.119150   
50%    455.500000   -0.214607   30.121215    8.000000    0.861596    1.637117   
75%    670.750000   -0.132954   39.719345   10.000000    0.941486    2.312282   
max    898.000000   -0.010924  115.700904   14.000000    1.000000    8.484189   

              mse        rmse  
count  854.000000  854.000000  
mean     6.955555    2.225946  
std     11.319743    1.415296  
min      0.000000    0.000000  
25%      1.823197    1.350258  
50%      3.904192    1.975902  
75%      7.430743    2.725938  
max     99.519188    9.975930
Out[37]:
Make this Notebook Trusted to load map: File -> Trust Notebook